Research on the Pattern Extraction of Academy of Classical Learning’s Buildings based on Image Enhancement Technology

Abstract

Conclusion This paper proposes a workflow for extracting architectural decorative patterns based on image enhancement technology and computer-aided design tools, including image enhancement and pattern standardization correction and extraction. The extraction results of typical pattern samples show that this workflow can batch enhance architectural decoration image materials, enhance their recognition and extract standard patterns. To a certain extent, this workflow can effectively reduce the equipment and material sampling environment requirements for the on-site acquisition of architectural decorative patterns and improve the availability of original materials. On this basis, geometric auxiliary lines can quickly locate the patterns in the image and reduce the difficulty of detail extraction. The features and innovations of this research are as follows: (1) A systematic, standardized, and efficient workflow based on computer-aided design technology is proposed for the problem of insufficient systematisms and low efficiency in the extraction of architectural decorative patterns. The innovative application of the CLAHE algorithm and adaptive bilateral filtering optimizes the problems of inefficiency and limited expertise produced by using PS to preprocess images. At the same time, it is proposed to use computer-aided design tools to draw geometric auxiliary lines for standardized correction and extraction of patterns, which significantly reduces the time cost and difficulty of pattern extraction. (2) This process simplifies the work of pattern extraction to drawing, cutting, and connecting simple geometric auxiliary lines, which breaks the limitation of professional ability. Even non-design researchers can quickly get started and perform pattern extraction. In addition, this pattern standardization extraction process is universal and applies to the digitization of all complex traditional patterns. present, this workflow still In research, researchers can try to realize automatic pattern extraction through edge recognition, curve fitting, machine learning, and other technologies based on the auxiliary line extraction method in paper.